Paracelsus' poison

Paracelsus' poison

Herbal Medicines: what do the labels say?

Herbal medicine labelling may not comply with their ARTG listing. Ian Musgrave

I’ve just spent over two days doing radio and TV interviews about a paper from my colleagues and I on labelling of herbal medicines.

It’s not actually physically printed yet but you can get an electronic version ahead of printed publication here if you are at an academic institution (and the press release here) if you are not. Our paper was noticed by Australian Doctor, and an article on it turned up in the Sydney Morning Herald and the Age.

So why all the interest (and some degree of concern)?

We found that in a survey of 121 herbal remedies advertised and sold on the internet from Australian sites that 18%, nearly one in five, were not on the Australian Register of Therapeutic Goods. Of the ones that were on the ARTG, overall 60% had active ingredients listed on their labels that did not match the active ingredients that permitted under their ARTG license. There was significant variation in concentrations of contents, or substitution of contents, or addition of active ingredients, or a combination of the above.

Note again that these were all Australian sites, either associated with manufactures, wholesalers or retailers with a physical presence in Australia. The Therapeutic Goods Administration does not regulate herbal medicines sold on the internet from overseas sites

We also went out and purchased a subset of these medicines from health food stores and pharmacies. 53% of these did not comply with their official ARTG listing.

Table. Comparison of the Website vs. TGA Ingredient Concentrations**

Arthritis Cold & Flu GI Stress PMS Total
Ingredients Match N % N % N % N % N % N %
Yes 10 37.0 7 25.0 5 41.7 8 34.8 0 0 30 30
No 12 44.4 18 64.3 7 58.3 13 56.5 9 100 59 60
N/A 5 18.5 3 10.7 0 0 2 8.7 0 0 10 10
Total 27 100 28 100 12 100 23 100 9 100 99 100
GI – gastrointestinal; PMS – premenstrual tension; N/A - unable to determine

Is this surprising?

No, this is consistent with what has been found in a previous small survey of complementary medicines in Australia by the TGA. This found that 90% of the randomly surveyed medicines were non-compliant with one or more of the regulations, and 71% had manufacturing or quality issues. Our larger survey focused exclusively on herbal medicines in specific therapeutic categories, but the general trend is similar.

It is also consistent with yet unpublished work we have done.

This is also consistent with a study of herbal medicines in Canada, which has a moderately similar regulatory regime.

So what is Australia’s regulatory regime?

In Australia herbal remedies which make therapeutic claims are treated as medicines, unlike some other countries, such as the USA, where they are treated as food supplements. There are some exceptions, for example some homeopathic medicines that have concentrations of non-biological active ingredients so low as to be unlikely to have any of the starting materials in them. I should note that we actually looked at more than 121 preparations, but several were exempt from ARTG listing, and not included.

However, in general, if a remedy or preparation makes a therapeutic claim and has substantial levels of active ingredients, it needs to be on the ARTG.

Within the ARTG, a remedy/medicine can be registered or listed. Registered medicines can make strong therapeutic claims, and need strong evidentiary backup. Typically they are supported by a variety of biological studies, clinical trials and safety and toxicological studies.

Listed medicines can make only general claims, such as “may assist with …”, “for support of…”. The evidentiary portfolio is much less stringent, and the sponsors are required to hold evidence that these remedies/are safe (although the TGA may not necessarily inspect this evidence), as well as ensuring that the materials in the remedy or medicine “only contain low risk ingredients in acceptable amounts”. Generally, listed medicines are considered low risk.

Most herbal remedies/medicines are listed.

The TGA listing process is largely trust-based; with sponsors self-certifying that the product meets TGA listing requirements. This requires post-marketing surveillance by the TGA. Again, a random audit by the TGA in 2009-2010 found 90% non-compliance of complementary medicines with one or more of the regulations.

One of the requirements of listing on the ARTG is that you cannot make substantial changes to the active ingredients post-approval. If this is done, then the medicine/remedy is no longer compliant, and a new ARTG lodgment must be made.

When you change the concentration of active ingredient by 5-10 fold, this represents a substantial change, as does substituting active ingredients and adding in new ones.

Is going from 5 to 50 mg of Ginger really going to be a worry?

Going from 5mg of Ginger to 50 mg of ginger may not be harmful, but it is the principle involved, if people cannot be bothered to pay attention to these regulations, what about the Good Manufacturing Protocol regulations? Going from 250 mg Ginseng to 2500 mg may not be harmful in itself (although serious side effects have been reported), but may increase the likelihood of interactions with conventional medicine.

As well there is the issue of substitution and addition. Substituting one herb for another has potential health implications. The concentration of active compounds in plants can vary significantly between cultivars, let alone species, and there is the risk that a plant with higher concentrations of active compound has been substituted, with the attendant increase in risk of adverse events or interactions with conventional medicines.

Adding in other active ingredients also has associated risks. A herbal manufacturer may think it’s a good idea to add wintergreen to their herbal anti-arthritis preparation. Wintergreen sounds nice and fresh and wholesome. But the active compounds in Wintergreen are a group of salicylates, related to aspirin. This has the potential to have adverse effects on its own (e.g. people who have allergies to salicylates or ulcers) or interact with other medications (e.g. low dose aspirin or warfarin where it could increase the chance of bleeding).

Reapplying for ARTG listing after such changes gives the TGA a chance to check the safety of these combinations, and ensure the labelling carries adequate warnings and information relating to the changed composition.

Thus the high level of non-compliance we found, following on from the high level of non-compliance found in the TGA audits, is cause for concern.

The Complementary Healthcare Council have responded to media reports (which, to be fair, were a little bit excitable) with this press release. I’d like to highlight this paragraph:

“The study authors failed to recognise that the expression of herbal extracts, as required on an ARTG entry, can be quantified differently than the expression on the label of the medicine or company website for the product. For example, the public summary on the ARTG may show the ‘input’ amount of the extract, while the label and website may show the ‘equivalent’ amount of the extract. Both forms of expression are correct and compliant with the regulations.”

At this very moment I am looking at a web page listing that shows the active compound as “Equivalent: substance redacted x g” while the TGA public summary shows “Equivalent: substance redacted z g”.

Contrary to what the CHC seems to think, we do know the regulations and can not only read but can also do maths. We did discuss concentrations with Regulatory Compliance Unit of the TGA, as well as sending them all our findings, before we went ahead and published.

The different units allowed were a great headache for us (we have a little whinge about this in our paper). We spent an instructive afternoon weighing herbal teabags to ensure they complied with the TGA listing (they did) and where we could not reconcile the TGA listing with the packet/website listing of a herbal medicine, it was put into the “unable to determine” category (N/A) rather than listed as non-compliant.

The CHC claim that “Australia is recognized as having one of the highest regarded and most tightly regulated systems in the world”. Yet that system relies critically on the cooperation of the herbal medicine sponsors. Our study implies, as does the previous TGA study, that this cooperation is lacking. The persistent intransigence of complementary medicine sponsors when asked to remove non-compliant medicines also speaks to this lack of cooperation (see here and here).

Of course, it would be ideal to actually measure the contents of these medicines, rather than rely on the labels, and we are doing such a study. Watch this space!